Back

Coordinated immune-epithelial dynamics in the nasal epithelium protect against respiratory virus infection

Yeo, Y. Y.; Chen, H.-Y.; Zhu, B.; Wang, Y.; Lee, I. T.; Nakayama, T.; Wang, Y.; Qiu, H.; Juergens, K. B.; Chung, S.; Yan, C. H.; Kummerer, C. S.; Michel, H. A.; Ma, A. Z.; Wu, W.; Nayak, J. V.; Nolan, G. P.; McIlwain, D. R.; Tzankov, A.; Matter, M. S.; Wu, C.-T.; Jiang, S.

2026-06-09 immunology
10.64898/2026.06.08.726938 bioRxiv
Show abstract

Epithelial remodeling is a hallmark of host antiviral responses that strengthen barrier defenses against infection. Our current understanding of viral pathophysiology within the nasal epithelium, the initial site of most respiratory viral infections, remains incomplete due to limited understanding of the native tissue architecture and protective epithelial remodeling during immune events. Leveraging coronavirus disease 2019 (COVID-19) as a model, we applied spatial multi-omics on nasal cross-sectional tissues to characterize the immune-epithelial landscape during infection, identifying a coordinated increase in goblet cells and suppressive macrophages associated with elevated IL13 in the immune compartment. Our results further reveal that goblet cell and suppressive macrophage enrichment are spatially linked in proximity to IL13-expressing CD4 T cells, consistent with IL13-driven remodeling in situ. Using a primary human nasal air-liquid interface model, we demonstrate that IL13 alone is sufficient to remodel epithelial composition and morphology, subsequently restricting viral infection by reshaping the apical mucus barrier of the nasal epithelium. Our findings uncover a spatially organized, IL13-driven circuit for immune-epithelial remodeling as a protective barrier against respiratory viral infections.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.